Identification device, identification method, and recording medium recording identification program
An identification device includes an inputter which receives image information of a person photographed by a camera, and a controller which identifies the person and detects parts, which are at least a head and hands, of the person based on the image information, thereby identifying a motion of the person based on the identified person, the detected parts, and a motion model in which a motion of a person is registered for every person, and outputs the identified motion of the person.
1. An identification device comprising:
a memory that stores instructions; and
a processor that, when executing the instructions stored in the memory, performs operations comprising:
receiving an image of a person photographed by a camera;
identifying the person based on the image;
detecting skeleton positions of parts, including at least a head and hands, of the person based on the image;
identifying a motion of the person; and
outputting the identified motion of the person,
wherein the processor, when executing the instructions stored in the memory, further performs operations comprising:
determining whether or not a skeleton detection is possible based on the detected skeleton positions,
when it is determined that the skeleton detection is possible, identifying the motion of the person based on the detected skeleton positions, and
when it is determined that the skeleton detection is not possible, identifying the motion of the person based on the identified person, the detected skeleton positions and a motion model in which a motion of each of a plurality of persons is registered.
2. The identification device according to claim 1 , wherein
the processor, when executing the instructions stored in the memory, further performs operations comprising:
generating a motion model based on the identified person, the detected skeleton positions, and the identified motion of the person, and
storing the generated motion model.
3. The identification device according to claim 1 , wherein
the processor, when executing the instructions stored in the memory, further performs operations comprising:
determining whether or not the motion model is available when it is determined that the skeleton detection is not possible, and
identifying the motion of the person based on the identified person, the detected skeleton positions, and the motion model and outputting the identified motion of the person, when it is determined that the motion model is available.
4. The identification device according to claim 1 , wherein
it is determined that the skeleton detection is possible when all of the skeleton positions are detected, and
it is determined that the skeleton detection is not possible when some of the skeleton positions are detected.
5. An identification method comprising:
receiving an image of a person photographed by a camera;
identifying the person based on the image;
detecting skeleton positions of parts, including at least a head and hands, of the person, based on the image;
identifying a motion of the person;
outputting the identified motion of the person,
wherein the identification method further comprising:
determining whether or not a skeleton detection is possible based on the detected skeleton positions,
when it is determined that the skeleton detection is possible, identifying the motion of the person based on the detected skeleton positions, and
when it is determined that the skeleton detection is not possible, identifying the motion of the person based on the identified person, the detected skeleton positions and a motion model in which a motion of each of a plurality of persons is registered.
6. A non-transitory computer readable recording medium recording an identification program which is executed in a computer, the computer identifying a motion of a person, and which makes the computer execute:
receiving an image of the person photographed by a camera;
identifying the person based on the image;
detecting skeleton positions of parts, including at least a head and hands, of the person based on the image;
identifying a motion of the person;
outputting the identified motion of the person,
wherein the identification program further makes the computer execute:
determining whether or not a skeleton detection is possible based on the detected skeleton positions,
when it is determined that the skeleton detection is possible, identifying the motion of the person based on the detected skeleton positions, and
when it is determined that the skeleton detection is not possible, identifying the motion of the person based on the identified person, the detected skeleton positions and a motion model in which a motion of each of a plurality of persons is registered.